aboutsummaryrefslogtreecommitdiffstats
path: root/contrib/libs/clang14-rt/lib/tsan/rtl/tsan_interface_atomic.cpp
blob: f794a2fcdd0df7fa372e017fa8a80fe333f10563 (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
//===-- tsan_interface_atomic.cpp -----------------------------------------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
//
// This file is a part of ThreadSanitizer (TSan), a race detector.
//
//===----------------------------------------------------------------------===//

// ThreadSanitizer atomic operations are based on C++11/C1x standards.
// For background see C++11 standard.  A slightly older, publicly
// available draft of the standard (not entirely up-to-date, but close enough
// for casual browsing) is available here:
// http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2011/n3242.pdf
// The following page contains more background information:
// http://www.hpl.hp.com/personal/Hans_Boehm/c++mm/

#include "sanitizer_common/sanitizer_placement_new.h"
#include "sanitizer_common/sanitizer_stacktrace.h"
#include "sanitizer_common/sanitizer_mutex.h"
#include "tsan_flags.h"
#include "tsan_interface.h"
#include "tsan_rtl.h"

using namespace __tsan;

#if !SANITIZER_GO && __TSAN_HAS_INT128
// Protects emulation of 128-bit atomic operations.
static StaticSpinMutex mutex128;
#endif

#if SANITIZER_DEBUG
static bool IsLoadOrder(morder mo) {
  return mo == mo_relaxed || mo == mo_consume
      || mo == mo_acquire || mo == mo_seq_cst;
}

static bool IsStoreOrder(morder mo) {
  return mo == mo_relaxed || mo == mo_release || mo == mo_seq_cst;
}
#endif

static bool IsReleaseOrder(morder mo) {
  return mo == mo_release || mo == mo_acq_rel || mo == mo_seq_cst;
}

static bool IsAcquireOrder(morder mo) {
  return mo == mo_consume || mo == mo_acquire
      || mo == mo_acq_rel || mo == mo_seq_cst;
}

static bool IsAcqRelOrder(morder mo) {
  return mo == mo_acq_rel || mo == mo_seq_cst;
}

template<typename T> T func_xchg(volatile T *v, T op) {
  T res = __sync_lock_test_and_set(v, op);
  // __sync_lock_test_and_set does not contain full barrier.
  __sync_synchronize();
  return res;
}

template<typename T> T func_add(volatile T *v, T op) {
  return __sync_fetch_and_add(v, op);
}

template<typename T> T func_sub(volatile T *v, T op) {
  return __sync_fetch_and_sub(v, op);
}

template<typename T> T func_and(volatile T *v, T op) {
  return __sync_fetch_and_and(v, op);
}

template<typename T> T func_or(volatile T *v, T op) {
  return __sync_fetch_and_or(v, op);
}

template<typename T> T func_xor(volatile T *v, T op) {
  return __sync_fetch_and_xor(v, op);
}

template<typename T> T func_nand(volatile T *v, T op) {
  // clang does not support __sync_fetch_and_nand.
  T cmp = *v;
  for (;;) {
    T newv = ~(cmp & op);
    T cur = __sync_val_compare_and_swap(v, cmp, newv);
    if (cmp == cur)
      return cmp;
    cmp = cur;
  }
}

template<typename T> T func_cas(volatile T *v, T cmp, T xch) {
  return __sync_val_compare_and_swap(v, cmp, xch);
}

// clang does not support 128-bit atomic ops.
// Atomic ops are executed under tsan internal mutex,
// here we assume that the atomic variables are not accessed
// from non-instrumented code.
#if !defined(__GCC_HAVE_SYNC_COMPARE_AND_SWAP_16) && !SANITIZER_GO \
    && __TSAN_HAS_INT128
a128 func_xchg(volatile a128 *v, a128 op) {
  SpinMutexLock lock(&mutex128);
  a128 cmp = *v;
  *v = op;
  return cmp;
}

a128 func_add(volatile a128 *v, a128 op) {
  SpinMutexLock lock(&mutex128);
  a128 cmp = *v;
  *v = cmp + op;
  return cmp;
}

a128 func_sub(volatile a128 *v, a128 op) {
  SpinMutexLock lock(&mutex128);
  a128 cmp = *v;
  *v = cmp - op;
  return cmp;
}

a128 func_and(volatile a128 *v, a128 op) {
  SpinMutexLock lock(&mutex128);
  a128 cmp = *v;
  *v = cmp & op;
  return cmp;
}

a128 func_or(volatile a128 *v, a128 op) {
  SpinMutexLock lock(&mutex128);
  a128 cmp = *v;
  *v = cmp | op;
  return cmp;
}

a128 func_xor(volatile a128 *v, a128 op) {
  SpinMutexLock lock(&mutex128);
  a128 cmp = *v;
  *v = cmp ^ op;
  return cmp;
}

a128 func_nand(volatile a128 *v, a128 op) {
  SpinMutexLock lock(&mutex128);
  a128 cmp = *v;
  *v = ~(cmp & op);
  return cmp;
}

a128 func_cas(volatile a128 *v, a128 cmp, a128 xch) {
  SpinMutexLock lock(&mutex128);
  a128 cur = *v;
  if (cur == cmp)
    *v = xch;
  return cur;
}
#endif

template <typename T>
static int AccessSize() {
  if (sizeof(T) <= 1)
    return 1;
  else if (sizeof(T) <= 2)
    return 2;
  else if (sizeof(T) <= 4)
    return 4;
  else
    return 8;
  // For 16-byte atomics we also use 8-byte memory access,
  // this leads to false negatives only in very obscure cases.
}

#if !SANITIZER_GO
static atomic_uint8_t *to_atomic(const volatile a8 *a) {
  return reinterpret_cast<atomic_uint8_t *>(const_cast<a8 *>(a));
}

static atomic_uint16_t *to_atomic(const volatile a16 *a) {
  return reinterpret_cast<atomic_uint16_t *>(const_cast<a16 *>(a));
}
#endif

static atomic_uint32_t *to_atomic(const volatile a32 *a) {
  return reinterpret_cast<atomic_uint32_t *>(const_cast<a32 *>(a));
}

static atomic_uint64_t *to_atomic(const volatile a64 *a) {
  return reinterpret_cast<atomic_uint64_t *>(const_cast<a64 *>(a));
}

static memory_order to_mo(morder mo) {
  switch (mo) {
  case mo_relaxed: return memory_order_relaxed;
  case mo_consume: return memory_order_consume;
  case mo_acquire: return memory_order_acquire;
  case mo_release: return memory_order_release;
  case mo_acq_rel: return memory_order_acq_rel;
  case mo_seq_cst: return memory_order_seq_cst;
  }
  DCHECK(0);
  return memory_order_seq_cst;
}

template<typename T>
static T NoTsanAtomicLoad(const volatile T *a, morder mo) {
  return atomic_load(to_atomic(a), to_mo(mo));
}

#if __TSAN_HAS_INT128 && !SANITIZER_GO
static a128 NoTsanAtomicLoad(const volatile a128 *a, morder mo) {
  SpinMutexLock lock(&mutex128);
  return *a;
}
#endif

template <typename T>
static T AtomicLoad(ThreadState *thr, uptr pc, const volatile T *a, morder mo) {
  DCHECK(IsLoadOrder(mo));
  // This fast-path is critical for performance.
  // Assume the access is atomic.
  if (!IsAcquireOrder(mo)) {
    MemoryAccess(thr, pc, (uptr)a, AccessSize<T>(),
                 kAccessRead | kAccessAtomic);
    return NoTsanAtomicLoad(a, mo);
  }
  // Don't create sync object if it does not exist yet. For example, an atomic
  // pointer is initialized to nullptr and then periodically acquire-loaded.
  T v = NoTsanAtomicLoad(a, mo);
  SyncVar *s = ctx->metamap.GetSyncIfExists((uptr)a);
  if (s) {
    SlotLocker locker(thr);
    ReadLock lock(&s->mtx);
    thr->clock.Acquire(s->clock);
    // Re-read under sync mutex because we need a consistent snapshot
    // of the value and the clock we acquire.
    v = NoTsanAtomicLoad(a, mo);
  }
  MemoryAccess(thr, pc, (uptr)a, AccessSize<T>(), kAccessRead | kAccessAtomic);
  return v;
}

template<typename T>
static void NoTsanAtomicStore(volatile T *a, T v, morder mo) {
  atomic_store(to_atomic(a), v, to_mo(mo));
}

#if __TSAN_HAS_INT128 && !SANITIZER_GO
static void NoTsanAtomicStore(volatile a128 *a, a128 v, morder mo) {
  SpinMutexLock lock(&mutex128);
  *a = v;
}
#endif

template <typename T>
static void AtomicStore(ThreadState *thr, uptr pc, volatile T *a, T v,
                        morder mo) {
  DCHECK(IsStoreOrder(mo));
  MemoryAccess(thr, pc, (uptr)a, AccessSize<T>(), kAccessWrite | kAccessAtomic);
  // This fast-path is critical for performance.
  // Assume the access is atomic.
  // Strictly saying even relaxed store cuts off release sequence,
  // so must reset the clock.
  if (!IsReleaseOrder(mo)) {
    NoTsanAtomicStore(a, v, mo);
    return;
  }
  SlotLocker locker(thr);
  {
    auto s = ctx->metamap.GetSyncOrCreate(thr, pc, (uptr)a, false);
    Lock lock(&s->mtx);
    thr->clock.ReleaseStore(&s->clock);
    NoTsanAtomicStore(a, v, mo);
  }
  IncrementEpoch(thr);
}

template <typename T, T (*F)(volatile T *v, T op)>
static T AtomicRMW(ThreadState *thr, uptr pc, volatile T *a, T v, morder mo) {
  MemoryAccess(thr, pc, (uptr)a, AccessSize<T>(), kAccessWrite | kAccessAtomic);
  if (LIKELY(mo == mo_relaxed))
    return F(a, v);
  SlotLocker locker(thr);
  {
    auto s = ctx->metamap.GetSyncOrCreate(thr, pc, (uptr)a, false);
    RWLock lock(&s->mtx, IsReleaseOrder(mo));
    if (IsAcqRelOrder(mo))
      thr->clock.ReleaseAcquire(&s->clock);
    else if (IsReleaseOrder(mo))
      thr->clock.Release(&s->clock);
    else if (IsAcquireOrder(mo))
      thr->clock.Acquire(s->clock);
    v = F(a, v);
  }
  if (IsReleaseOrder(mo))
    IncrementEpoch(thr);
  return v;
}

template<typename T>
static T NoTsanAtomicExchange(volatile T *a, T v, morder mo) {
  return func_xchg(a, v);
}

template<typename T>
static T NoTsanAtomicFetchAdd(volatile T *a, T v, morder mo) {
  return func_add(a, v);
}

template<typename T>
static T NoTsanAtomicFetchSub(volatile T *a, T v, morder mo) {
  return func_sub(a, v);
}

template<typename T>
static T NoTsanAtomicFetchAnd(volatile T *a, T v, morder mo) {
  return func_and(a, v);
}

template<typename T>
static T NoTsanAtomicFetchOr(volatile T *a, T v, morder mo) {
  return func_or(a, v);
}

template<typename T>
static T NoTsanAtomicFetchXor(volatile T *a, T v, morder mo) {
  return func_xor(a, v);
}

template<typename T>
static T NoTsanAtomicFetchNand(volatile T *a, T v, morder mo) {
  return func_nand(a, v);
}

template<typename T>
static T AtomicExchange(ThreadState *thr, uptr pc, volatile T *a, T v,
    morder mo) {
  return AtomicRMW<T, func_xchg>(thr, pc, a, v, mo);
}

template<typename T>
static T AtomicFetchAdd(ThreadState *thr, uptr pc, volatile T *a, T v,
    morder mo) {
  return AtomicRMW<T, func_add>(thr, pc, a, v, mo);
}

template<typename T>
static T AtomicFetchSub(ThreadState *thr, uptr pc, volatile T *a, T v,
    morder mo) {
  return AtomicRMW<T, func_sub>(thr, pc, a, v, mo);
}

template<typename T>
static T AtomicFetchAnd(ThreadState *thr, uptr pc, volatile T *a, T v,
    morder mo) {
  return AtomicRMW<T, func_and>(thr, pc, a, v, mo);
}

template<typename T>
static T AtomicFetchOr(ThreadState *thr, uptr pc, volatile T *a, T v,
    morder mo) {
  return AtomicRMW<T, func_or>(thr, pc, a, v, mo);
}

template<typename T>
static T AtomicFetchXor(ThreadState *thr, uptr pc, volatile T *a, T v,
    morder mo) {
  return AtomicRMW<T, func_xor>(thr, pc, a, v, mo);
}

template<typename T>
static T AtomicFetchNand(ThreadState *thr, uptr pc, volatile T *a, T v,
    morder mo) {
  return AtomicRMW<T, func_nand>(thr, pc, a, v, mo);
}

template<typename T>
static bool NoTsanAtomicCAS(volatile T *a, T *c, T v, morder mo, morder fmo) {
  return atomic_compare_exchange_strong(to_atomic(a), c, v, to_mo(mo));
}

#if __TSAN_HAS_INT128
static bool NoTsanAtomicCAS(volatile a128 *a, a128 *c, a128 v,
    morder mo, morder fmo) {
  a128 old = *c;
  a128 cur = func_cas(a, old, v);
  if (cur == old)
    return true;
  *c = cur;
  return false;
}
#endif

template<typename T>
static T NoTsanAtomicCAS(volatile T *a, T c, T v, morder mo, morder fmo) {
  NoTsanAtomicCAS(a, &c, v, mo, fmo);
  return c;
}

template <typename T>
static bool AtomicCAS(ThreadState *thr, uptr pc, volatile T *a, T *c, T v,
                      morder mo, morder fmo) {
  // 31.7.2.18: "The failure argument shall not be memory_order_release
  // nor memory_order_acq_rel". LLVM (2021-05) fallbacks to Monotonic
  // (mo_relaxed) when those are used.
  DCHECK(IsLoadOrder(fmo));

  MemoryAccess(thr, pc, (uptr)a, AccessSize<T>(), kAccessWrite | kAccessAtomic);
  if (LIKELY(mo == mo_relaxed && fmo == mo_relaxed)) {
    T cc = *c;
    T pr = func_cas(a, cc, v);
    if (pr == cc)
      return true;
    *c = pr;
    return false;
  }
  SlotLocker locker(thr);
  bool release = IsReleaseOrder(mo);
  bool success;
  {
    auto s = ctx->metamap.GetSyncOrCreate(thr, pc, (uptr)a, false);
    RWLock lock(&s->mtx, release);
    T cc = *c;
    T pr = func_cas(a, cc, v);
    success = pr == cc;
    if (!success) {
      *c = pr;
      mo = fmo;
    }
    if (success && IsAcqRelOrder(mo))
      thr->clock.ReleaseAcquire(&s->clock);
    else if (success && IsReleaseOrder(mo))
      thr->clock.Release(&s->clock);
    else if (IsAcquireOrder(mo))
      thr->clock.Acquire(s->clock);
  }
  if (success && release)
    IncrementEpoch(thr);
  return success;
}

template<typename T>
static T AtomicCAS(ThreadState *thr, uptr pc,
    volatile T *a, T c, T v, morder mo, morder fmo) {
  AtomicCAS(thr, pc, a, &c, v, mo, fmo);
  return c;
}

#if !SANITIZER_GO
static void NoTsanAtomicFence(morder mo) {
  __sync_synchronize();
}

static void AtomicFence(ThreadState *thr, uptr pc, morder mo) {
  // FIXME(dvyukov): not implemented.
  __sync_synchronize();
}
#endif

// Interface functions follow.
#if !SANITIZER_GO

// C/C++

static morder convert_morder(morder mo) {
  if (flags()->force_seq_cst_atomics)
    return (morder)mo_seq_cst;

  // Filter out additional memory order flags:
  // MEMMODEL_SYNC        = 1 << 15
  // __ATOMIC_HLE_ACQUIRE = 1 << 16
  // __ATOMIC_HLE_RELEASE = 1 << 17
  //
  // HLE is an optimization, and we pretend that elision always fails.
  // MEMMODEL_SYNC is used when lowering __sync_ atomics,
  // since we use __sync_ atomics for actual atomic operations,
  // we can safely ignore it as well. It also subtly affects semantics,
  // but we don't model the difference.
  return (morder)(mo & 0x7fff);
}

#  define ATOMIC_IMPL(func, ...)                                \
    ThreadState *const thr = cur_thread();                      \
    ProcessPendingSignals(thr);                                 \
    if (UNLIKELY(thr->ignore_sync || thr->ignore_interceptors)) \
      return NoTsanAtomic##func(__VA_ARGS__);                   \
    mo = convert_morder(mo);                                    \
    return Atomic##func(thr, GET_CALLER_PC(), __VA_ARGS__);

extern "C" {
SANITIZER_INTERFACE_ATTRIBUTE
a8 __tsan_atomic8_load(const volatile a8 *a, morder mo) {
  ATOMIC_IMPL(Load, a, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a16 __tsan_atomic16_load(const volatile a16 *a, morder mo) {
  ATOMIC_IMPL(Load, a, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a32 __tsan_atomic32_load(const volatile a32 *a, morder mo) {
  ATOMIC_IMPL(Load, a, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a64 __tsan_atomic64_load(const volatile a64 *a, morder mo) {
  ATOMIC_IMPL(Load, a, mo);
}

#if __TSAN_HAS_INT128
SANITIZER_INTERFACE_ATTRIBUTE
a128 __tsan_atomic128_load(const volatile a128 *a, morder mo) {
  ATOMIC_IMPL(Load, a, mo);
}
#endif

SANITIZER_INTERFACE_ATTRIBUTE
void __tsan_atomic8_store(volatile a8 *a, a8 v, morder mo) {
  ATOMIC_IMPL(Store, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
void __tsan_atomic16_store(volatile a16 *a, a16 v, morder mo) {
  ATOMIC_IMPL(Store, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
void __tsan_atomic32_store(volatile a32 *a, a32 v, morder mo) {
  ATOMIC_IMPL(Store, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
void __tsan_atomic64_store(volatile a64 *a, a64 v, morder mo) {
  ATOMIC_IMPL(Store, a, v, mo);
}

#if __TSAN_HAS_INT128
SANITIZER_INTERFACE_ATTRIBUTE
void __tsan_atomic128_store(volatile a128 *a, a128 v, morder mo) {
  ATOMIC_IMPL(Store, a, v, mo);
}
#endif

SANITIZER_INTERFACE_ATTRIBUTE
a8 __tsan_atomic8_exchange(volatile a8 *a, a8 v, morder mo) {
  ATOMIC_IMPL(Exchange, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a16 __tsan_atomic16_exchange(volatile a16 *a, a16 v, morder mo) {
  ATOMIC_IMPL(Exchange, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a32 __tsan_atomic32_exchange(volatile a32 *a, a32 v, morder mo) {
  ATOMIC_IMPL(Exchange, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a64 __tsan_atomic64_exchange(volatile a64 *a, a64 v, morder mo) {
  ATOMIC_IMPL(Exchange, a, v, mo);
}

#if __TSAN_HAS_INT128
SANITIZER_INTERFACE_ATTRIBUTE
a128 __tsan_atomic128_exchange(volatile a128 *a, a128 v, morder mo) {
  ATOMIC_IMPL(Exchange, a, v, mo);
}
#endif

SANITIZER_INTERFACE_ATTRIBUTE
a8 __tsan_atomic8_fetch_add(volatile a8 *a, a8 v, morder mo) {
  ATOMIC_IMPL(FetchAdd, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a16 __tsan_atomic16_fetch_add(volatile a16 *a, a16 v, morder mo) {
  ATOMIC_IMPL(FetchAdd, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a32 __tsan_atomic32_fetch_add(volatile a32 *a, a32 v, morder mo) {
  ATOMIC_IMPL(FetchAdd, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a64 __tsan_atomic64_fetch_add(volatile a64 *a, a64 v, morder mo) {
  ATOMIC_IMPL(FetchAdd, a, v, mo);
}

#if __TSAN_HAS_INT128
SANITIZER_INTERFACE_ATTRIBUTE
a128 __tsan_atomic128_fetch_add(volatile a128 *a, a128 v, morder mo) {
  ATOMIC_IMPL(FetchAdd, a, v, mo);
}
#endif

SANITIZER_INTERFACE_ATTRIBUTE
a8 __tsan_atomic8_fetch_sub(volatile a8 *a, a8 v, morder mo) {
  ATOMIC_IMPL(FetchSub, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a16 __tsan_atomic16_fetch_sub(volatile a16 *a, a16 v, morder mo) {
  ATOMIC_IMPL(FetchSub, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a32 __tsan_atomic32_fetch_sub(volatile a32 *a, a32 v, morder mo) {
  ATOMIC_IMPL(FetchSub, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a64 __tsan_atomic64_fetch_sub(volatile a64 *a, a64 v, morder mo) {
  ATOMIC_IMPL(FetchSub, a, v, mo);
}

#if __TSAN_HAS_INT128
SANITIZER_INTERFACE_ATTRIBUTE
a128 __tsan_atomic128_fetch_sub(volatile a128 *a, a128 v, morder mo) {
  ATOMIC_IMPL(FetchSub, a, v, mo);
}
#endif

SANITIZER_INTERFACE_ATTRIBUTE
a8 __tsan_atomic8_fetch_and(volatile a8 *a, a8 v, morder mo) {
  ATOMIC_IMPL(FetchAnd, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a16 __tsan_atomic16_fetch_and(volatile a16 *a, a16 v, morder mo) {
  ATOMIC_IMPL(FetchAnd, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a32 __tsan_atomic32_fetch_and(volatile a32 *a, a32 v, morder mo) {
  ATOMIC_IMPL(FetchAnd, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a64 __tsan_atomic64_fetch_and(volatile a64 *a, a64 v, morder mo) {
  ATOMIC_IMPL(FetchAnd, a, v, mo);
}

#if __TSAN_HAS_INT128
SANITIZER_INTERFACE_ATTRIBUTE
a128 __tsan_atomic128_fetch_and(volatile a128 *a, a128 v, morder mo) {
  ATOMIC_IMPL(FetchAnd, a, v, mo);
}
#endif

SANITIZER_INTERFACE_ATTRIBUTE
a8 __tsan_atomic8_fetch_or(volatile a8 *a, a8 v, morder mo) {
  ATOMIC_IMPL(FetchOr, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a16 __tsan_atomic16_fetch_or(volatile a16 *a, a16 v, morder mo) {
  ATOMIC_IMPL(FetchOr, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a32 __tsan_atomic32_fetch_or(volatile a32 *a, a32 v, morder mo) {
  ATOMIC_IMPL(FetchOr, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a64 __tsan_atomic64_fetch_or(volatile a64 *a, a64 v, morder mo) {
  ATOMIC_IMPL(FetchOr, a, v, mo);
}

#if __TSAN_HAS_INT128
SANITIZER_INTERFACE_ATTRIBUTE
a128 __tsan_atomic128_fetch_or(volatile a128 *a, a128 v, morder mo) {
  ATOMIC_IMPL(FetchOr, a, v, mo);
}
#endif

SANITIZER_INTERFACE_ATTRIBUTE
a8 __tsan_atomic8_fetch_xor(volatile a8 *a, a8 v, morder mo) {
  ATOMIC_IMPL(FetchXor, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a16 __tsan_atomic16_fetch_xor(volatile a16 *a, a16 v, morder mo) {
  ATOMIC_IMPL(FetchXor, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a32 __tsan_atomic32_fetch_xor(volatile a32 *a, a32 v, morder mo) {
  ATOMIC_IMPL(FetchXor, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a64 __tsan_atomic64_fetch_xor(volatile a64 *a, a64 v, morder mo) {
  ATOMIC_IMPL(FetchXor, a, v, mo);
}

#if __TSAN_HAS_INT128
SANITIZER_INTERFACE_ATTRIBUTE
a128 __tsan_atomic128_fetch_xor(volatile a128 *a, a128 v, morder mo) {
  ATOMIC_IMPL(FetchXor, a, v, mo);
}
#endif

SANITIZER_INTERFACE_ATTRIBUTE
a8 __tsan_atomic8_fetch_nand(volatile a8 *a, a8 v, morder mo) {
  ATOMIC_IMPL(FetchNand, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a16 __tsan_atomic16_fetch_nand(volatile a16 *a, a16 v, morder mo) {
  ATOMIC_IMPL(FetchNand, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a32 __tsan_atomic32_fetch_nand(volatile a32 *a, a32 v, morder mo) {
  ATOMIC_IMPL(FetchNand, a, v, mo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a64 __tsan_atomic64_fetch_nand(volatile a64 *a, a64 v, morder mo) {
  ATOMIC_IMPL(FetchNand, a, v, mo);
}

#if __TSAN_HAS_INT128
SANITIZER_INTERFACE_ATTRIBUTE
a128 __tsan_atomic128_fetch_nand(volatile a128 *a, a128 v, morder mo) {
  ATOMIC_IMPL(FetchNand, a, v, mo);
}
#endif

SANITIZER_INTERFACE_ATTRIBUTE
int __tsan_atomic8_compare_exchange_strong(volatile a8 *a, a8 *c, a8 v,
    morder mo, morder fmo) {
  ATOMIC_IMPL(CAS, a, c, v, mo, fmo);
}

SANITIZER_INTERFACE_ATTRIBUTE
int __tsan_atomic16_compare_exchange_strong(volatile a16 *a, a16 *c, a16 v,
    morder mo, morder fmo) {
  ATOMIC_IMPL(CAS, a, c, v, mo, fmo);
}

SANITIZER_INTERFACE_ATTRIBUTE
int __tsan_atomic32_compare_exchange_strong(volatile a32 *a, a32 *c, a32 v,
    morder mo, morder fmo) {
  ATOMIC_IMPL(CAS, a, c, v, mo, fmo);
}

SANITIZER_INTERFACE_ATTRIBUTE
int __tsan_atomic64_compare_exchange_strong(volatile a64 *a, a64 *c, a64 v,
    morder mo, morder fmo) {
  ATOMIC_IMPL(CAS, a, c, v, mo, fmo);
}

#if __TSAN_HAS_INT128
SANITIZER_INTERFACE_ATTRIBUTE
int __tsan_atomic128_compare_exchange_strong(volatile a128 *a, a128 *c, a128 v,
    morder mo, morder fmo) {
  ATOMIC_IMPL(CAS, a, c, v, mo, fmo);
}
#endif

SANITIZER_INTERFACE_ATTRIBUTE
int __tsan_atomic8_compare_exchange_weak(volatile a8 *a, a8 *c, a8 v,
    morder mo, morder fmo) {
  ATOMIC_IMPL(CAS, a, c, v, mo, fmo);
}

SANITIZER_INTERFACE_ATTRIBUTE
int __tsan_atomic16_compare_exchange_weak(volatile a16 *a, a16 *c, a16 v,
    morder mo, morder fmo) {
  ATOMIC_IMPL(CAS, a, c, v, mo, fmo);
}

SANITIZER_INTERFACE_ATTRIBUTE
int __tsan_atomic32_compare_exchange_weak(volatile a32 *a, a32 *c, a32 v,
    morder mo, morder fmo) {
  ATOMIC_IMPL(CAS, a, c, v, mo, fmo);
}

SANITIZER_INTERFACE_ATTRIBUTE
int __tsan_atomic64_compare_exchange_weak(volatile a64 *a, a64 *c, a64 v,
    morder mo, morder fmo) {
  ATOMIC_IMPL(CAS, a, c, v, mo, fmo);
}

#if __TSAN_HAS_INT128
SANITIZER_INTERFACE_ATTRIBUTE
int __tsan_atomic128_compare_exchange_weak(volatile a128 *a, a128 *c, a128 v,
    morder mo, morder fmo) {
  ATOMIC_IMPL(CAS, a, c, v, mo, fmo);
}
#endif

SANITIZER_INTERFACE_ATTRIBUTE
a8 __tsan_atomic8_compare_exchange_val(volatile a8 *a, a8 c, a8 v,
    morder mo, morder fmo) {
  ATOMIC_IMPL(CAS, a, c, v, mo, fmo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a16 __tsan_atomic16_compare_exchange_val(volatile a16 *a, a16 c, a16 v,
    morder mo, morder fmo) {
  ATOMIC_IMPL(CAS, a, c, v, mo, fmo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a32 __tsan_atomic32_compare_exchange_val(volatile a32 *a, a32 c, a32 v,
    morder mo, morder fmo) {
  ATOMIC_IMPL(CAS, a, c, v, mo, fmo);
}

SANITIZER_INTERFACE_ATTRIBUTE
a64 __tsan_atomic64_compare_exchange_val(volatile a64 *a, a64 c, a64 v,
    morder mo, morder fmo) {
  ATOMIC_IMPL(CAS, a, c, v, mo, fmo);
}

#if __TSAN_HAS_INT128
SANITIZER_INTERFACE_ATTRIBUTE
a128 __tsan_atomic128_compare_exchange_val(volatile a128 *a, a128 c, a128 v,
    morder mo, morder fmo) {
  ATOMIC_IMPL(CAS, a, c, v, mo, fmo);
}
#endif

SANITIZER_INTERFACE_ATTRIBUTE
void __tsan_atomic_thread_fence(morder mo) { ATOMIC_IMPL(Fence, mo); }

SANITIZER_INTERFACE_ATTRIBUTE
void __tsan_atomic_signal_fence(morder mo) {
}
}  // extern "C"

#else  // #if !SANITIZER_GO

// Go

#  define ATOMIC(func, ...)               \
    if (thr->ignore_sync) {               \
      NoTsanAtomic##func(__VA_ARGS__);    \
    } else {                              \
      FuncEntry(thr, cpc);                \
      Atomic##func(thr, pc, __VA_ARGS__); \
      FuncExit(thr);                      \
    }

#  define ATOMIC_RET(func, ret, ...)              \
    if (thr->ignore_sync) {                       \
      (ret) = NoTsanAtomic##func(__VA_ARGS__);    \
    } else {                                      \
      FuncEntry(thr, cpc);                        \
      (ret) = Atomic##func(thr, pc, __VA_ARGS__); \
      FuncExit(thr);                              \
    }

extern "C" {
SANITIZER_INTERFACE_ATTRIBUTE
void __tsan_go_atomic32_load(ThreadState *thr, uptr cpc, uptr pc, u8 *a) {
  ATOMIC_RET(Load, *(a32*)(a+8), *(a32**)a, mo_acquire);
}

SANITIZER_INTERFACE_ATTRIBUTE
void __tsan_go_atomic64_load(ThreadState *thr, uptr cpc, uptr pc, u8 *a) {
  ATOMIC_RET(Load, *(a64*)(a+8), *(a64**)a, mo_acquire);
}

SANITIZER_INTERFACE_ATTRIBUTE
void __tsan_go_atomic32_store(ThreadState *thr, uptr cpc, uptr pc, u8 *a) {
  ATOMIC(Store, *(a32**)a, *(a32*)(a+8), mo_release);
}

SANITIZER_INTERFACE_ATTRIBUTE
void __tsan_go_atomic64_store(ThreadState *thr, uptr cpc, uptr pc, u8 *a) {
  ATOMIC(Store, *(a64**)a, *(a64*)(a+8), mo_release);
}

SANITIZER_INTERFACE_ATTRIBUTE
void __tsan_go_atomic32_fetch_add(ThreadState *thr, uptr cpc, uptr pc, u8 *a) {
  ATOMIC_RET(FetchAdd, *(a32*)(a+16), *(a32**)a, *(a32*)(a+8), mo_acq_rel);
}

SANITIZER_INTERFACE_ATTRIBUTE
void __tsan_go_atomic64_fetch_add(ThreadState *thr, uptr cpc, uptr pc, u8 *a) {
  ATOMIC_RET(FetchAdd, *(a64*)(a+16), *(a64**)a, *(a64*)(a+8), mo_acq_rel);
}

SANITIZER_INTERFACE_ATTRIBUTE
void __tsan_go_atomic32_exchange(ThreadState *thr, uptr cpc, uptr pc, u8 *a) {
  ATOMIC_RET(Exchange, *(a32*)(a+16), *(a32**)a, *(a32*)(a+8), mo_acq_rel);
}

SANITIZER_INTERFACE_ATTRIBUTE
void __tsan_go_atomic64_exchange(ThreadState *thr, uptr cpc, uptr pc, u8 *a) {
  ATOMIC_RET(Exchange, *(a64*)(a+16), *(a64**)a, *(a64*)(a+8), mo_acq_rel);
}

SANITIZER_INTERFACE_ATTRIBUTE
void __tsan_go_atomic32_compare_exchange(
    ThreadState *thr, uptr cpc, uptr pc, u8 *a) {
  a32 cur = 0;
  a32 cmp = *(a32*)(a+8);
  ATOMIC_RET(CAS, cur, *(a32**)a, cmp, *(a32*)(a+12), mo_acq_rel, mo_acquire);
  *(bool*)(a+16) = (cur == cmp);
}

SANITIZER_INTERFACE_ATTRIBUTE
void __tsan_go_atomic64_compare_exchange(
    ThreadState *thr, uptr cpc, uptr pc, u8 *a) {
  a64 cur = 0;
  a64 cmp = *(a64*)(a+8);
  ATOMIC_RET(CAS, cur, *(a64**)a, cmp, *(a64*)(a+16), mo_acq_rel, mo_acquire);
  *(bool*)(a+24) = (cur == cmp);
}
}  // extern "C"
#endif  // #if !SANITIZER_GO